Role of EUS combined with a newly modified scoring system to detect pancreatic high-grade precancerous lesions.

Abstract

Although pancreatic cancer (PC) has an extremely poor prognosis, the 5-year survival rate of patients with pancreatic high-grade precancerous lesion without invasive carcinoma (PHP) is favorable. PHP diagnosis and identification of patients requiring intervention are needed. We aimed to validate a modified PC detection scoring system regarding its detection ability for PHP and PC in the general population. We modified an existing PC detection scoring system that incorporates low-grade risk (LGR) factors (family history, presence of diabetes mellitus [DM] or worsening DM, heavy drinking, smoking, stomach symptoms, weight loss, and pancreatic enzyme) and high-grade risk (HGR) factors (new-onset DM, familial PC, jaundice, tumor biomarkers, chronic pancreatitis, intraductal papillary mucinous neoplasm, cysts, hereditary PC syndrome, and hereditary pancreatitis). Each factor was scored as one point; LGR score ≥3 points and/or HGR score ≥1 point (positive scores) were indicative of PC. The newly modified scoring system incorporated main pancreatic duct dilation as an HGR factor. The PHP diagnosis rate using this scoring system combined with EUS was prospectively analyzed. Among 544 patients with positive scores, 10 had PHP. The diagnosis rates were 1.8% for PHP and 4.2% for invasive PC. Although the number of LGR and HGR factors tended to increase with PC progression, none of the individual factors were significantly different between patients with PHP and those without lesions. The newly modified scoring system evaluating multiple factors associated with PC could potentially identify patients with higher risk of PHP or PC.